Abstract: A series of mutual citation networks of Chinese S&T Core Journals grouped by subjects are described in matrix form according to CSTPCD 2010. Furthermore, normalized matrices of similarity coefficients are imported in Pajek, the popular visualization tool, to map those mutual citation networks. Based on the analysis about those figures, three basic patterns including Star-Form, Tree-Form, Cluster-Form and two integrative patterns including Multicenter-Form and Bus-Form are distinguished and defined. What we could learn from the study about patterns of mutual citation networks of journals grouped by subjects is to indicate intuitively the characteristics of subjects or fields, to show the relation and statement of academic exchanging between one journal and others in subject, to monitor subjects’ evolving, merging and classifying, to support the running of S&T Journals Confidence Index and to discover and avoid some academic misconduct such as untrue mutual citations from a cluster of journals.